This competing continuation application of 5U18DA07673-04 requests 3 years of funding to complete an ongoing multi-site study examining the efficacy of individual cognitive therapy (CT), supportive-expressive (SE) psychodynamic therapy, and individual drug counseling (IDC) when combined with group drug counseling compared with group drug counseling (GDC) alone for 480 patients who meet DSM-lV criteria or cocaine dependence. This project has been funded as a Cooperative Agreement involving two NIDA collaborators, a Coordinating Center at the U. of Pennsylvania, and treatment sites at Western Psychiatric Institute and Clinic (Pittsburgh), Brookside Hospital (Nashua, NH), McLean and Mass General Hospital (Boston), and U. of Pennsylvania. During the last grant period we have recruited and trained our research teams, developed efficient screening procedures, standardized treatment methods that can be administered with a high degree of specificity and fidelity. These goals were accomplished in the course of conducting pilot study involving 314 randomized patients that provided preliminary evidence for our hypotheses. The aim trial began in July 1994 and thus far we have recruited 129 patients. We anticipate recruiting a total of 288 patients between July 1994 and the beginning of the competing continuation in September 1995 and an additional 192 patients in the first year of the renewal. Specifically, we hypothesize that CT, SE, and IDC will be better than GDC; that for patients with high levels of psychiatric severity CT and SE will be better than lDC; and that for patients with an external coping style CT will be better than SE treatments consists of a 6-month active phase followed by 3 month booster phase. Assessments ar one monthly through the active phase and at 9, 1 2, 1 5, and 1 8 months post randomization to examine he long-term effects of treatment. The main outcome measures include urine results (done weekly), addiction Severity Index Drug Use composite, and self-report of cocaine use. A variety of secondary outcome measures are also obtained. The data analysis will use both cross- sectional and longitudinal approaches.